Boldine與Laurolitsine衍生物之製備與抑制黃嘌呤氧化酶活性之研究

Abstract

Boldine為具有Aporphine架構之生物鹼，主要分離自杯軸花科(Monimiaceae)植物波多樹(Boldo, Peumus boldus Molina)之葉與樹皮。Laurolitsine亦為具有Aporphine架構之新型生物鹼，其主要由樟科(Lauraceae)植物台灣雅楠(Phoebe formosana)之莖部分離出來。根據文獻資料可得知此兩化合物均具有多種生物活性作用。 首先，根據電腦分子模擬分析發現：對於黃嘌呤氧化酶(xanthine oxidase)，Aporphine與Secoaporphine呈現良好的對接(嵌合)結果。然而，在實際的體外活性測試結果中，此類化合物對於黃嘌呤氧化酶的抑制效果卻不如預期。原因可能為Aporphine中的氮原子會自動質子化，造成Aporphine變成游離鹼(free base)而帶有正電荷，使得在電腦分子模擬的對接(嵌合)系統中會增加額外的離子交互作用。然而，在生物體或藥理活性測試中的酸鹼值大約在pH 7.5，Aporphine在此環境下由於酸性不足以使氮原子質子化，故在電腦分子模擬與體外活性測試中之結果產生出入。為了驗證此推論，本論文合成具有四級胺或三級胺之Aporphine與Secoaporphine，並進行體外的活性測試加以證實。 以Boldine (1)作為起始物，先進行N-methylation，加熱時溫度的高低會產生不同的化合物。於常溫，產生N-Methylated quaternary ammonium salt；於高溫，則產生開環的N-Methylsecoboldine (5)。然而，直接常溫下進行N-methylation產生N-Methylboldinium halide (4)之產率小於30%，故改為先將Boldine (1)經由乙醯基保護再進行N-methylation，最後以酸水解反應切掉乙醯基，即可得到較高產率之N-Methylboldinium halide (4)。另外，亦將N-Methylsecoboldine (5)先經過乙醯化取得N-Methyldiacetylsecoboldine (6)後，再進行N-methylation與酸水解反應，得到N,N-Dimethylsecoboldinium chloride (7)。 此外，亦透過電腦分子模擬設計與分析其他Aporphine之相關結構，以期許發現其他具有抑制黃嘌呤氧化酶之成分。此部分之主要來源為修飾過去於O-methylation製備Glaucine時所得到反應不完全的Monomethyl Derivatives of Boldine以及透過半合成之方式將對接(docking)結果優良之化學結構製備出來，並進一步探討這類化合物與黃嘌呤氧化酶之活性關係。 在修飾Monomethyl Derivatives of Boldine的部分，以N-Methyllauroletenine (A)與Predicentrine (B)之混合物為起始物。由於兩者之物理特性相近，難以分離純化，故由先前發現之2-Hydroxyaporphine可藉由溶劑熱解反應(Solvolysis reaction)將Predicentrine (B)進行開環，得到Secopredicentrine (8)。此時，混合物N-Methyllauroletenine (A)與Predicentrine (B)之極性差異變大，因此便可將Secopredicentrine (8)分離純化，並進一步將Secopredicentrine (8)進行環化，得到2a-Methyl-7-O-methyllitebamine (9)。再者，亦透過電腦分子模擬設計出兩種系列具有三個苯環及一個六圓環之化學結構，兩系列主要差異於：系列一為環化之二級胺，系列二則為環化之三級胺，即其氮原子上帶有一個甲基。系列一以原油的Laurolitsine作為起始物，經簡易半合成方法可大量製備N-Acetyllaurolitsine (10)，接著進行開環與水解反應。然而，根據不同的反應條件，可得到開環但未水解之N-Acetylsecolaurolitsine (11)與開環且水解之Secolaurolitsine (12)。接著，將Secolaurolitsine (12)進行曼尼希反應(Mannich reaction)，即可得到環化之Norlitebamine (13)。系列二則再次以Boldine (1)作為起始物，經簡易半合成方法可大量製備Secoboldine (14)，並同樣進行曼尼希反應(Mannich reaction)以得到相關之環化衍生物 (15, 16)。 在活性測試結果中，可證實第一部分之氮原子的自動質子化會造成電腦計算與實際藥理活性測試的結果產生出入。於第二部分與第三部分所製備出的化合物亦進行活性的探討。整體而言，可知N-Dimethylsecoboldinium chloride (7)具有最佳的抑制能力，而N-Methylsecoboldine (5)、2a-Methyl-7-O-methyllitebamine (9)、Norlitebamine (13)、2a-Methyllitebamine (15)以及2a-Ethyllitebamine (16)亦具有一定的抑制效果。以上七種化合物對於黃嘌呤氧化酶的抑制效果雖然不盡相同，仍期許未來這些化合物能有助於Aporphine類之高尿酸血症或痛風用藥之開發。

Boldine is an alkaloid with skeleton of Aporphine. It is mainly separated from leaves and barks of Boldo, Peumus boldus Molina which belongs to Monimiaceae family. Laurolitsine is also a new alkaloid with skeleton of Aporphine. And it is mainly separated from stems of Phoebe formosana which belongs to Lauraceae family. According to literature, there are many kinds of biological activities in these two alkaloids. First, according to the computer molecular modeling analysis, Aporphine and Secoaporphine demonstrate high docking scores on xanthine oxidase, even much better than Febuxostat which is the clinically used drug nowadays. Nevertheless, some kinds of Aporphines and Secoaporphines displayed only medium inhibitory activities in vitro assay. The reason might be attributed to the auto-protonation of N-containing Aporphines and Secoaporphines in docking system. The condition makes the free base Aporphines positive charged and arising else ionic interaction. So that there were different results between in vitro assay and computer molecular modeling. To confirm the hypothesis above, Aporphines and Secoaporphines with quaternary or tertiary amines have been semisynthesized in this thesis. The starting material Boldine (1) was converted to N-methylation. The temperature played a key role in this condition. In room temperature, The N-methylated quaternary ammonium salt was prepared; however, it produced N-Methylsecoboldine (5) in high temperature. Still, the yield of N-Methylboldinium halide (4) was less than 30% in this method. Following subsequently with three reaction steps on Boldine (1), protection with the acetyl group, N-methylation and hydrolysis, the N-Methylated quaternary ammonium salt was prepared in higher yield. Moreover, N-Methylsecoboldine (5) was also converted to acetylation, N-methylation and hydrolysis, and N,N-Dimethylsecoboldinium salt (7) was prepared. Besides, hoping to find another structures that are also able to inhibit xanthine oxidase, there are some relative structures to Aporphine are designed and analyzed by computer molecular modeling analysis. These parts are to modify the Monomethyl Derivatives of Boldine from past work and to semisynthesize the compounds with good docking scores. In modification of Monomethyl Derivatives of Boldine, it is starting from the mixture of N-Methyllauroletenine (A) and Predicentrine (B). Owing to the similar physical characteristics, these two compounds are separated with difficulty. Thus, to get ring-opening Secopredicentrine (8) from Predicentrine (B) by Solvolysis reaction of 2-Hydroxyaporphine. The polarities are different from N-Methyllauroletenine (A) and Secopredicentrine (8). Then, Secopredicentrine (8) can be separated easily from the mixture and also form 2a-Methyl-7-O-methyllitebamine (9) by cyclization. Moreover, there are two series compounds with three benzene rings and a six-membered ring by computer molecular modeling design and analysis. Series-1 is secondary amine, and Series-2 is tertiary amine. In Series-1, starting from crude Laurolitsine, N-Acetyllaurolitsine (10) was prepared by a facile semisynthetic method. According to different conditions, N-Acetylsecolaurolitsine (11) and Secolaurolitsine (12) were prepared. Then, Secolaurolitsine (12) formed Norlitebamine (13) via Mannich reaction. In Series-2, starting from Boldine (1) again, Secoboldine (14) was prepared by a facile semisynthetic method. Then, Secoboldine (14) formed relative derivatives (15, 16) via Mannich reaction as well. In the result of bioassay, it is provable that the auto-protonation of N-containing Aporphine makes the result of computer molecular modeling different from the result of in vitro assay in part one. And the compounds prepared in part two and part three also do the bioassay of xanthine oxidase. In general, N-Dimethylsecoboldinium chloride (7) is with the best inhibition against xanthine oxidase. And N-Methylsecoboldine (5), 2a-Methyl-7-O-methyllitebamine (9), Norlitebamine (13), 2a-Methyllitebamine (15) and 2a-Ethyllitebamine (16) are with not bad inhibition against xanthine oxidase as well. These results will serve as a vital index for developing potential drugs of the Aporphine type for helping treat hyperuricemia or gout.